BioTIMEr (version 0.3.0)

getLinearRegressions: Get Linear Regressions BioTIME

Description

Fits linear regression models to getAlphaMetrics or getBetaMetrics outputs

Usage

getLinearRegressions(x, pThreshold = 0.05)

Value

Returns a single long data.frame with results of linear regressions (slope, p-value, significance, intercept) for each assemblageID.

Arguments

x

(data.frame) BioTIME data table in the format of the output of getAlphaMetrics or getBetaMetrics functions

pThreshold

(numeric) P-value threshold for statistical significance

Details

The function getLinearRegression fits simple linear regression models (see lm for details) for a given output ('data') of either getAlphaMetrics or getBetaMetrics function. The typical model has the form metric ~ year. Note that assemblages with less than 3 time points and/or single species time series are removed.

Examples

  x <- data.frame(
    resamp = 1L,
    YEAR = rep(rep(2010:2015, each = 4), times = 4),
    Species = c(replicate(n = 8L * 6L, sample(letters[1L:10L], 4L, replace = FALSE))),
    ABUNDANCE = rpois(24 * 8, 10),
    assemblageID = rep(LETTERS[1L:8L], each = 24)
  )
  alpham <- getAlphaMetrics(x, "ABUNDANCE")
  getLinearRegressions(x = alpham, pThreshold = 0.01) |> head(10)

  betam <- getBetaMetrics(x = x, "ABUNDANCE")
  getLinearRegressions(x = betam) |> head(10)